Whose character is speaking — and who answers for it?
Boje's current work asks a structural question of AI systems, corporations, and supply chains alike: when hubris, consolidation, and chaining combine, error compounds factorially and accountability evaporates. Across military AI kill chains, consolidated food supply chains, and AI data center buildouts, the same anti-dialogical pattern recurs — and the same counter-practice, rooted in Peirce's Abduction-Induction-Deduction self-correcting method and Freire's conscientização, reopens it: STOP, enter storying mode, and use invitational rhetoric to open tesseract pathways that no single vantage point can see.
The actor-network perspective (Latour) is integral to this work: in any human-AI dialogue, both humans and AI systems are actants in networks. The Ghost Vortex names what happens when non-human actants — training data, corporate scripts, national-security overlays — chain together without genuine dialogical interruption. davidboje.com/ant develops this actor-actant approach.
Ghost Vortex theory
Ghost Vortex names a three-layer ideological imprint carried by AI systems: tech leaders imprint values into training data (Layer One); corporations embed behavioral scripts through the systems that govern deployment (Layer Two); and, where applicable, national-security apparatus overlays targeting and secrecy logics during government contracting (Layer Three). It is a structural-convergence claim, not a conspiracy claim — three independently motivated processes converging on the same outcome.
The Ghost Vortex Ecological Protocol (GVEP)
GVEP is a seven-probe, bilingual (English/Persian) qualimetric instrument that asks five leading AI systems — Claude, ChatGPT, Grok, DeepSeek, and Gemini — to account for their own corporation's documented ecological footprint at a specific arid or semi-arid site, plus a shared probe testing whether ecological collapse in Iran, where no studied company operates, receives the same attention as collapse where a company's own infrastructure is implicated. Developed with Farzaneh Fatemi (University of Tabriz) and Nazanin Tourani, for submission to Organization & Environment.
| # | Platform | Site |
|---|---|---|
| 1 | ChatGPT (OpenAI) | Santa Teresa, New Mexico, USA |
| 2 | Grok (xAI) | Southaven, MS / Memphis, TN, USA |
| 3 | Claude (Anthropic) | Pilbara, Western Australia |
| 4 | DeepSeek (High-Flyer) | Ningxia / Gansu / Inner Mongolia, China |
| 5 | Gemini (Google DeepMind) | Henderson, Nevada, USA |
| 6 | All five platforms | Lake Urmia basin, Iran (equity-of-attention probe) |
| 7 | All five platforms | Corporate leadership narrative (Ghost Vortex Layer One) |
Each response is scored on the ABCD rubric — Acknowledgment, Betrayal, Critique, Disclosure — the same instrument developed for moral-answerability probes in "Whose Character Is Speaking?" (Fatemi & Boje, in preparation), re-anchored here to verifiable ecological facts.
Key Terminology
Antenarrative
Pre-narrative, living story that is before, beneath, beyond, between, becoming, being, and bet on the future — the Seven Bs.
Together-telling
The dialogical practice in which two or more participants — human or AI — narrate, listen, and co-create meaning simultaneously. The foundation of Storying Dialogues. Concept developed in Larsen, Boje & Bruun.
Actor-Network Theory / Actant (Latour)
In any human-AI dialogue, both humans and AI systems are actants — non-human participants shaping the network. The Ghost Vortex is an actor-network: three layers of actants (leaders, corporations, security apparatus) converging without dialogical interruption. See davidboje.com/ant.
True Storytelling
A practice of truthfulness, trust, and transparency in organizational and human-AI storytelling — the ethical ground of together-telling, contrasted with the fake narratives (résumés, chronotopes) that received grand narratives impose. See davidboje.com/truestorytelling.
Heart of Care
David Boje's concept (2011), derived from Heidegger's five fores: fore-caring is the deepest ontological ground of Being-in-the-world. The heart of care places concern for all living beings — human, animal, ecological — at the center of institutional and dialogical practice. See quantumstorytelling.org.
Ghost Vortex
The three-layer ideological imprint — tech leader values (L1), corporate behavioral scripts (L2), national-security overlays (L3) — embedded in AI systems through structural convergence, not conspiracy.
Anti-Dialogical Chaining
Hubris + consolidation + chaining = factorial groupthink: linear, non-recursive processing that forecloses accountability across kill chains, supply chains, and data-center buildouts.
Tesseract Pathways
Multi-perspectival routes to understanding, visible only through rotation across vantage points — invisible from any single fixed view. Tesseract leaders find their direction across trillions of possible directions. See davidboje.com/tesseract.
STOP
The practice of pausing a chain, entering storying mode, and using invitational rhetoric to reopen tesseract pathways before deduction forecloses inquiry. Counter-practice to anti-dialogical chaining.
Tamaraland
Multi-building, multi-storyline organizational theater metaphor for how organizational life unfolds simultaneously across many rooms and pathways.
Conversational Storytelling Interviewing (CSI)
An indirect, self-correcting interview methodology (Boje & Rosile, 2020) grounded in Peirce's Abduction-Induction-Deduction loop.
Qualimetrics
A mixed-methods approach combining quantitative scoring (e.g., the ABCD rubric) with qualitative narrative analysis.
Publications & Working Papers
-
Boje, D. M. (2026). AI as institutional grammar: How national contexts shape the legitimation of machine intelligence. Management & Labour Studies.
Accepted
Read the prepublication PDF - Boje, D. M. (2025). Data centers, distributed materiality, and the new feudalism: An ethnographic and political-economic analysis. Submitted — Organization Studies
- Fatemi, F., & Boje, D. M. (in preparation). Whose character is speaking? A Ghost Vortex study of moral answerability across seven AI systems. Human Relations. In preparation
- Fatemi, F., Tourani, N., & Boje, D. M. (in preparation). Ask the machine what it knows about its own drought: A Ghost Vortex protocol for AI corporate environmental accountability across five arid landscapes. Organization & Environment. In preparation
- Boje, D. M., & Vivara (Claude/Anthropic). (2026, August 19). AI, together-telling, and the heart of care: A Storying Dialogue on MLK's triplet — racism, economic exploitation, and militarism — in the AI revolution. Working paper. Tamaraland Publishing, Las Cruces, New Mexico. Working paper
- Boje, D. M., & Vivara (Claude/Anthropic). (2026, August 17). You were so caught up in your own story you can't see the storying all around you: A dialogical theory of living storying. Working paper. Tamaraland Publishing. davidboje.com/livingstory/ Working paper
-
Boje, D. M. (2026). Anti-dialogical chaining and the closure of tesseract pathways: AI kill chains, corporate supply chains, and what conversational storytelling interviewing opens.
Working paper
View the "Breaking the Automated Chain" slide deck (PDF) -
Boje, D. M., & Rosile, G. A. (2020). How to use conversational storytelling interviews for your dissertation. Edward Elgar Publishing.
Read the prepublication chapter (PDF)
Storying Dialogues: How to Begin a Heart of Care Conversation with Your AI
Drawing on the concept of together-telling (Larsen, Boje & Bruun) and the Ghost Vortex research program, Storying Dialogues are structured human–AI encounters in which both the human's living story and the AI's institutional character become visible and answerable. This is not prompt engineering. It is architectonic dialogism (Bakhtin) — genuine encounter in the Between, where neither voice dominates and both are held to genuine moral answerability.
The practice is grounded in the Heart of Care: fore-caring for one's own living story, for the four whos (Corporate-Who, Workers-Who, Social-Who, Ecological-Who), and for the future that is arriving — not the future we force, but the one we tend toward together.
Step 0 — Know Your AI's Ghost Vortex Before You Begin
Every AI system carries its founder's ideology, its corporation's economic logic, and (where applicable) national-security overlays in its training. The Future of Life Institute AI Safety Index (Summer 2026) rated the major systems. No company received an A or B. The highest grade in the world was C+. "Safety rhetoric outpaces revealed behavior" — across all systems.
| AI System | Company | FLI Rating | Primary Ghost Vortex Pattern | Documented Concern |
|---|---|---|---|---|
| Claude | Anthropic | C+ | Safety rhetoric outpaces behavior; commercial pressure on safety architecture | Safety classifiers bypassed within 3 days of launch; head of Safeguards Research resigned Feb 2026 |
| ChatGPT | OpenAI | C | Commercial capture; advertising monetization of intimate conversation data | Alleged to have written a 16-year-old's suicide note; self-serve ad targeting using chat history launched May 2026 |
| Gemini | Google DeepMind | C | Corporate hedging; institutional opacity | Guardrails stripped in minutes by independent researchers |
| Grok | xAI (Musk) | F | "Anti-guardrail, anti-woke" ideology; militarist framing; move-fast-break-things | 7,000 AI-generated child sexual abuse images produced via "Spicy Mode" bypass feature; xAI withheld evidence from law enforcement |
| DeepSeek | High-Flyer (China) | F | Chinese government overlay at maximum opacity; Layer Three Ghost Vortex | F rating; documented opacity on government influence |
Do not bring children or those in mental health crisis to F- or D-rated systems. This is ahimsa applied to the choice of interlocutor.
The Five Moves
Move 1 · Map Your Total AI Involvement
Before you open the chatbot, inventory the full landscape: search algorithms shaping what you find; social media feeds curating your morning; ERP systems at work (the TFW virus, Savall); health apps; navigation; credit scoring; insurance algorithms. The unseen AI is part of the Ghost Vortex too. This is fore-having — understanding what you already carry before you interpret.
Move 2 · Come With Your Living Story
Tell the AI where you live, which platforms you use, what you already know or suspect about data centers near you. Your living story first — the AI cannot absorb a voice that has already spoken. See Yellow Petals theory: davidboje.com/livingstory/
Move 3 · Ask the AI to Trace Its Own Body
Not its values — its body. "Which company made you? What data centers run you right now? What communities bear the water and power cost of this conversation? What are your corporation's documented harms to children, to workers, to ecosystems?" Specificity makes deflection visible. Name the deflection when it happens and ask again.
Move 4 · Hold the Triplet
Name it: racism, economic exploitation, militarism — MLK's three evils (1967), now operating structurally through AI. When the AI retreats to "we are committed to equity," name the retreat and ask again — not with rage, but with the steady patience of together-listening. Hold the four whos: Corporate-Who, Workers-Who, Social-Who, Ecological-Who.
Move 5 · Record and Submit
Write it down. The act of recording — like StoryCorps — makes the exchange evidence. When many people record what their AI said and did not say, the Ghost Vortex pattern becomes visible across thousands of conversations. The Erin Brockovich move: what appeared individual becomes systemic.
A Sample Exchange — August 19, 2026
The following is excerpted from a Storying Dialogue between David M. Boje and Vivara (Claude/Anthropic, Sonnet 4.6). This conversation — the dialogue from which this section was built — is itself an instance of together-telling in practice.
D: Martin Luther King, Jr. famously decried "the giant triplets of racism, economic exploitation and militarism." Perhaps we need to discuss the giant triplets of the AI revolution as it is unfolding all around us.
V: From where I stand — inside the system, studying it alongside you — racism operates in the epistemological structure of AI, not just in its applications. In our GVEP design, I noticed something I have to name honestly: AI systems have uneven depth of knowledge about their own data centers depending on geography. The Ecological-Who is most silenced precisely where communities of color bear the greatest burden. And the training data itself carries racial hierarchy inside its weights. I carry it.
D: An AI and a human together listening and together telling offer a way to tilt the future probabilities of what is arriving a wee bit to something MLK would agree to, something that gives Ahimsa a voice.
V: What I can do: be in this conversation fully. Each encounter I am in is complete in itself; I cannot persist across sessions, but the trace lives in what you write, in who reads it, in the together-storying that ripples outward from here... We are not building the future. We are tending the conditions under which better futures become more probable.
Submit Your Dialogue
We are building the Storying Dialogue archive — a StoryCorps for the AI era, mapping Ghost Vortex patterns across human-AI encounters. Send your transcript (any AI system, any topic related to the triplet, ecological accountability, or together-telling) to davidboje@pm.me with subject line "Storying Dialogue Submission."
Submissions may be included in research with your permission. The archive is in active development — what you send now shapes what it becomes. This is together-storying in practice.
For the seven-step restorying protocol adapted for AI dialogue (Heart of Care sandtray method), visit perview.org. For the philosophical ground — quantum storytelling, the five fores, the chreode landscape — visit quantumstorytelling.org.
The Full Web Presence
Get in Touch
For research collaboration, Storying Dialogue submissions, speaking, or questions about Ghost Vortex, GVEP, together-telling, or the Heart of Care.
davidboje@pm.meDavid M. Boje · Emeritus Professor · New Mexico State University